Risk factors for hyperglycemia in COVID-19 patients treated with remdesivir

Area of Science:

• * Infectious Diseases
• * Pharmacology
• * Medical Informatics

Background:

• * Remdesivir is a key antiviral medication used in treating hospitalized patients with coronavirus disease 2019 (COVID-19).
• * Hyperglycemic adverse events (AEs) are potential complications associated with remdesivir administration.
• * Understanding the predictors of these AEs is essential for optimizing patient care and treatment strategies.

Purpose of the Study:

• * To identify factors associated with hyperglycemic AEs in COVID-19 patients receiving remdesivir.
• * To develop and evaluate machine learning models for predicting hyperglycemia-related complications.
• * To aid in the development of personalized remdesivir treatment regimens.

Main Methods:

• * A retrospective analysis of 1262 hospitalized COVID-19 patients treated with remdesivir.
• * Logistic regression analysis to assess the relationship between covariates and hyperglycemic AEs.
• * Application of machine learning algorithms (multivariate logistic regression, elastic net, random forest) for risk prediction.

Main Results:

• * Elevated body mass index (BMI) ≥23 kg/m², use of proton pump inhibitors, cholinergic medications, and presence of cardiovascular diseases were significantly associated with increased risk of hyperglycemic AEs.
• * Odds ratios for these risk factors ranged from approximately 1.97 to 2.73.
• * Machine learning models showed moderate predictive performance, with Area Under the Receiver Operating Characteristic Curve (AUROC) values ranging from 0.60 to 0.66.

Conclusions:

• * Specific patient characteristics and concomitant medications are linked to higher risks of hyperglycemia during remdesivir therapy for COVID-19.
• * Machine learning approaches show potential in developing risk stratification tools for these adverse events.
• * Findings support the need for tailored remdesivir treatment strategies, considering individual patient risk profiles.